RF Amplifiers Solid State vs IOTs vs Klystrons

RF Amplifiers: Solid State vs IOTs vs Klystrons Paco Sanchez, RF Group, CELLS (Consortium for the Exploitation of the Synchrotron Light Laboratory) Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 1

RF Amplifiers: Solid State vs IOTs vs Klystrons Acknowledgements : I want to thank to all of you, RF experts, for your presentations, from which I took plenty of data and advice. ALS (LBNL): Ken Baptiste ANL: Doug Horan, Ali Nassiri BESSY: Thomas Westphal BROOKHAVEN: Jim Rose CERN: Luca Arnaudon, Pablo Martínez, Eric Montesinos, Hans Frischholz DESY: Wolfhard Merz, Rüdiger Onken, Michael Ebert DIAMOND: Morten Jensen ELETTRA: Cristina Pasotti, Alessandro Fabris ESRF: Michel Langlois, Jean-Maurice Mercier, Jörn Jacob SLAC: Chris Adolphsen SOLEIL: Robert Lopes, Ti Ruan, Patrick Marchand SPRING-8: Shigeki Sasaki THALES: A. J. Durand ALBA: Francis Pérez and all our RF&Diagnostics team Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 2

RF Amplifiers: Solid State vs IOTs vs Klystrons RF Amplifiers 1. Klystrons 2. IOTs 3. Solid State Amplifiers 4. Some Criteria 5. Implementation 6. Discussions / Conclusions? Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 3

RF Amplifiers: Solid State vs IOTs vs Klystrons 1. RF Amplifiers: Klystrons Klystron principle: - High electron DC current produced in the klystron gun. - Current velocity modulated in the first cavity by the driving power. - The e-beam is then grouped in pulses, which are enhanced in three intermediate cavities (shorter pulses) - This pulsed stream induces a high field in the last cavity, out of which the high power is decoupled. Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 4

RF Amplifiers: Solid State vs IOTs vs Klystrons 1. RF Amplifiers: Klystrons Klystron performances: Freq. (ANKA): 500 MHz Output Power: 250 k. W Efficiency: 62% Gain: > 40 d. B Freq. (CERN): 400 MHz Output Power: 300 k. W Efficiency: 62% Gain: > 35 d. B ANKA Klystron (e 2 V) Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 5

RF Amplifiers: Solid State vs IOTs vs Klystrons 1. RF Amplifiers: Klystrons @ 400 MHz 1, 5 m @ 3000 MHz 4, 5 m Dimensions: Cavity size and drift tube length are related to wavelength The size of the interaction structure increases as the frequency decreases Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 6

RF Amplifiers: Solid State vs IOTs vs Klystrons 1. RF Amplifiers: Klystrons Output power and efficiency: Po = η Vk. Ib , * As η < 1 , increase of Pout requires increase of both beam power and cathode voltage (both related to space charge). Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 7

RF Amplifiers: Solid State vs IOTs vs Klystrons 1. RF Amplifiers: Klystrons Output power and efficiency: * Pout increases with Vk , but the increase of Vk is limited by breakdown in the gun , i. e. vacuum breakdown * Breakdown occurs when the electric field in the range 5 x 103 to 104 k. V/mm * High Output Power ( up to a few MWcw in the frequency range, required in particle accelerators) * High Efficiency ( up to about 70%, by inserting the intermediate cavities): One tuned to the second harmonic, the others to frequencies well above, ensuring that they are inductive at the operating frequency * High efficiency is only achieved at rated output power, decreases rapidly at lower power levels. Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 8

RF Amplifiers: Solid State vs IOTs vs Klystrons 1. RF Amplifiers: Klystrons Gain, Lifetime, experience: * High gain, typically 40 d. B are obtained at an output bandwidth of a few percent * Price per W of output power decreases by increasing the output power per unit. * Very sturdy, especially when cavities are an integral part of the klystron body. (klystrons with an output power of about 100 k. W and more) * TV klystrons (< 60 k. W) are equipped with external cavities. More fragile due to RF contacts body-cavities, and to the many ceramics in the klystron body * Experience: most frequently failure is related to the electron gun (heater or HV feedthrough, heater filament interrupt, high voltage arcing, HV ceramic leak) * For a long cathode life (> 100000 h): the emission density should be < 1 A/cm 2. * In high power klystrons the low cathode loading is achieved by the convergence from cathode to beam cross-section area, i. e. the cathode surface is about 10 to 30 times that of the electron beam. Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 9

RF Amplifiers: Solid State vs IOTs vs Klystrons 2. RF Amplifiers: IOTs An IOT is a hybrid creature RF OUT RF IN PLANAR TRIODE IOT Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 KLYSTRON 10

RF Amplifiers: Solid State vs IOTs vs Klystrons 2. RF Amplifiers: IOTs * From the triode, the IOT inherit a cathode and a grid. Its beam is not velocity modulated, as in a klystron, but density modulated as in a triode. It means a RF voltage swing must exist between cathode and grid. Contrary to the klystron case, the cathode (and the grid) are included in a RF circuit. * From the klystron, the IOT gets nearly everything else: • Linear beam • Pierce gun • Hollow anode • Magnetic field for focussing • TM 010 output cavity • Collector Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 11

RF Amplifiers: Solid State vs IOTs vs Klystrons 2. RF Amplifiers: IOTs In the ALBA case: PRF = 80 k. W VHV = - 36 k. V IHV = 3. 2 A h = 71 % A RF OUT + _ H. V supply e. g. -36 k. V / 3. 2 A Since: PRFin = 266 W Gain = 25 d. B Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 Grid supply e. g. 150 V / 250 m. A - + RF IN + Heater supply e. g. 12 V / 25 A 12

RF Amplifiers: Solid State vs IOTs vs Klystrons 2. RF Amplifiers: IOTs Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 13

RF Amplifiers: Solid State vs IOTs vs Klystrons 2. RF Amplifiers: IOTs Ø Tube Modifications (THALES ED): 1. Bigger output window diameter 2. Conical insulator between grid anode Ø Advantages 1. Better suited for High Power cw operation 2. Electric Field in the ceramic insulator decreased 3. Losses also decreased 4. Less stray electrons impinge on the tube Ø Drawback 1. Dedicated product for monofrequency applications Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 80 k. W version (TV) 90 k. W version (LS) 14

RF Amplifiers: Solid State vs IOTs vs Klystrons 2. RF Amplifiers: IOTs Ø Output Cavities Modifications (TED-CELLS): 1. Enlarge coax output coupling from 4 -1/16” to 61/8” 2. Redesigned primary cavity to house the new tube 90 k. W 3. Increase the secondary output cavity depth 4. Add a post (resonance vs. harmonics) 5. Second arc detector Ø Ø Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 Advantages 1. Cope with the higher cw power operation 2. Simplified tuning devices in primary cavity Drawback 15

RF Amplifiers: Solid State vs IOTs vs Klystrons 2. RF Amplifiers: IOTs Ø Ø Advantage 1. Prevent thermal frequency shifts Ø Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 Other Modifications (TEDCELLS): 1. Modify accordingly the cavities coupling loops: • Output loop • Prim-sec cavity loop 2. Water cooling added: • Primary cavity • Inter-cavity loop Drawback 1. Add water cooling circuits in the RF Tx 16

RF Amplifiers: Solid State vs IOTs vs Klystrons 2. RF Amplifiers: IOTs Ø Cavity Simulations (CELLS): 1. 2. CST-Microwave Stutio ™ Output circuit simulated to obtain the cavities’ dimensions and the coupling loops’ dimensions 3. Material conductivity adjusted in order to get the beam voltage at 90 k. W IOT Conductive material in the drift tube gap to simulate the beam 4. Primary output circuit 5. Bandwidth evaluation from the S 11 parameter Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 Output coupling Prim- sec coupling Secondary output circuit 17

RF Amplifiers: Solid State vs IOTs vs Klystrons 2. RF Amplifiers: IOTs Ø Computed BW compared to the measured one: 4. 7 MHz Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 6. 8 MHz 18

RF Amplifiers: Solid State vs IOTs vs Klystrons Ø 2. RF Amplifiers: IOTs Duration test, at 80 k. W, during more than 24 h, at THALES premises Ø Also tested in factory at 90 k. W during ½ hour Ø Harmonics at Dir-Coupler (6” coax output): Ø Bandwidth = 6. 8 MHz; and Tuning capabilities = ± 5 MHz Ø Problems: IOT SN 499413 Ionic Pump reconditioning in factory Ø IOT SN 499443 Poor efficiency reconditioned on site (FIL on 1 week) Ø IOT SN 611024 High Body current reconditioned on site (FIL on 3 days) Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 19

RF Amplifiers: Solid State vs IOTs vs Klystrons 2. RF Amplifiers: IOTs New problem appeared on May 04, 12 h 00 (yesterday, during lunchtime) IOT SN 611024 ceramic broke (to be investigated) Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 20

RF Amplifiers: Solid State vs IOTs vs Klystrons 2. RF Amplifiers: IOTs The 2 IOTs already high power tested on site (out of 14 in future operation) IOT TH 793 22, 7 d. B Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 21

RF Amplifiers: Solid State vs IOTs vs Klystrons 2. RF Amplifiers: IOTs IOT TH 793 67 % Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 22

RF Amplifiers: Solid State vs IOTs vs Klystrons 2. RF Amplifiers: IOTs Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 23

RF Amplifiers: Solid State vs IOTs vs Klystrons 2. RF Amplifiers: IOTs Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 24

RF Amplifiers: Solid State vs IOTs vs Klystrons 2. RF Amplifiers: IOTs From the previous table: Average Life (Fail) Average (In _Use) CPI (1 / 1 ) 10842 h 17222 h e 2 V (1 / 10 ) 1219 h 7289 h TH 793 (14 / 6 ) 5834 h 9559 h TH 793 -1 (1 / 4 ) 4052 h 1109 h Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 25

RF Amplifiers: Solid State vs IOTs vs Klystrons 3. RF Amplifiers: SSA Alternative (e. g. SOLEIL): Solid State Technology Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 26

RF Amplifiers: Solid State vs IOTs vs Klystrons 3. RF Amplifiers: SSA Alternative in some labs: Solid State Technology Water Cooling Dissipater 8 -Way Splitter Amplifier Module 20 k. W Combiner 40 k. W Combiner 2 -Way Splitter 40 k. W Combiner Coaxial Output 35 k. W Directional Coupler 2. 5 k. W Combiner DC/DC Converter 2. 5 k. W Groups Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 27

RF Amplifiers: Solid State vs IOTs vs Klystrons 3. RF Amplifiers: SSA Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 28

RF Amplifiers: Solid State vs IOTs vs Klystrons 3. RF Amplifiers: SSA In the SOLEIL case: PRF = 315 W modules VDC = 50 V f = 352 MHz h = 62 % Gain = 13 d. B PRF = 400 W modules VDC = 50 V f = 476 MHz PRF = 700 W modules VDC = 50 V f = 352 MHz h ~ 70 % Gain ~ 20 d. B h > 70 % Gain > 20 d. B PRF = 180 k. W “TOWER” h ~ 60 % Gain ~ 53 d. B Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 29

RF Amplifiers: Solid State vs IOTs vs Klystrons 3. RF Amplifiers: SSA Running hours: Booster RF: 35 k. W SSA, with no trips after 20000 hours SR RF: 4 towers of 180 k. W at 352 MHz, with no impact in the operation. A 1: running time 20. 000 h (2006 -2009) and around 4% trans. failure A 2: running time 20. 000 h (2006 -2009) and around 1. 2% trans. failure A 3: running time 10. 000 h (2008 -2009) and around 1. 7% trans. failure A 4: running time 10. 000 h (2008 -2009) and around 1. 7% trans. failure Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 30

RF Amplifiers: Solid State vs IOTs vs Klystrons 4. RF Amplifiers: Some Criteria Running Hours: Klystron 1 MW (APS) 53000 h (5 units average) IOT 80 k. W (ALBA) 520 h (ups!) SSA 35 k. W, 20000 h (Booster) 250 k. W (CERN) 50000 h 80 k. W (ELETTRA) 5000 h (5 units TV type) 2500 h (2 units LS type) 180 k. W, 20000 h (SR x 2 towers) 180 k. W, 10000 h (SR x 2 towers) 60 k. W (ELETTRA) 26000 h (4 units average) . 80 k. W (DIAMOND) 4300 h (11 IOT average life on failure) 15000 h & 17000 h (oldest 2 IOTs) 40 k. W (ALS) 17500 h (1 unit, 40 k. W pk AM) 80 k. W (BESSY) 11300 h (1 unit) Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 31

RF Amplifiers: Solid State vs IOTs vs Klystrons 4. RF Amplifiers: Some Criteria Output Power, Efficiency and Gain: Klystron IOT 60 - 1000 k. W 80 k. W 700 W 180 k. W 62 % 71 % 70 % 60 % 40 d. B 25 d. B 20 d. B 53 d. B (module) (Tower) Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 SSA . 32

RF Amplifiers: Solid State vs IOTs vs Klystrons 4. RF Amplifiers: Some Criteria Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 33

RF Amplifiers: Solid State vs IOTs vs Klystrons 4. RF Amplifiers: Some Criteria Purchase Cost in k. Euro/ k. W: Cost Klystron 400 MHz / 250 k. W Klystron ~ 230 k€ ~ 0. 9 k€ / k. W IOTs 500 MHz / 80 k. W IOTs ~ 50 - 70 k€ ~ 0. 6 - 0. 9 k€ / k. W *SSA 352 MHz / 200 k. W SSA ~ 740 k€ ~ 3. 7 k€/ k. W (300 W) ~ 2. 8 k€/ k. W (700 W) *N. B: SSA from SOLEIL data (4 “towers” of 200 k. W). The complete transmitter includes transformer + HV rectifier + control system, …), but SOLEIL manpower is not included. All components were built in the industry according to the SOLEIL specifications) Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 34

RF Amplifiers: Solid State vs IOTs vs Klystrons 4. RF Amplifiers: Some Criteria Since SSA include power supplies, we have to add: ALBA IOTs (HVPS - PSM) 160 k. W ~ 230 k€ ~ 1. 44 k€/k. W PSM DESY Klystron (thyristor converter) 3040 k. W ~ 880 k€ ~ 0. 29 k€/k. W converter At 1. 3 GHz (SLAC needs 30 k. W): Purchase Cost: Cost Klystron ~ 500 k$ ~ 16. 7 k$ / k. W ~ 12. 7 k€ / k. W Klystron IOTs ~ 300 k$ IOTs ~ 10 k$ / k. W ~ 7. 6 k€ / k. W SSA ~ 480 k$ SSA ~ 16 k$ / k. W ~ 12. 2 k€ / k. W Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 35

RF Amplifiers: Solid State vs IOTs vs Klystrons 5. IMPLEMENTATION Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 36

RF Amplifiers: Solid State vs IOTs vs Klystrons 6. DISCUSSIONS / CONCLUSIONS? * Hard to collect data for comparison * Difficulties to get prices * SSA includes power supplies and development * IOTs including several modifications but… * New IOTs without long term data * Discussion is served! Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 37

RF Amplifiers: Solid State vs IOTs vs Klystrons MANY THANKS again to ALL of YOU Paco Sanchez - 06 th CWRF Workshop - ALBA, 04 -07 May 2010 38